Endotoxin Tolerance Disrupts Chromatin Remodeling and NF-B Transactivation at the IL-1 Promoter1
http://www.100md.com
免疫学杂志 2005年第13期
Abstract
The NF-B family plays a crucial role in the pathogenesis of highly lethal septicemia by modulating transcription of many innate and adaptive immunity genes. Two phases of NF-B activation occur: cytosolic activation and nuclear transactivation. Septicemia with multiorgan failure is associated with chronic activation of cytosolic NF-B with translocation and accumulation of increased levels of nuclear p65 in blood leukocytes. Paradoxically, NF-B-dependent transcription of many proinflammatory genes responding to bacterial LPS endotoxin (LPS) is persistently repressed during septicemia; this phenomenon of LPS tolerance is associated with immunosuppression and poor prognosis. This report suggests an explanation for this paradox. Using an in vitro human leukocyte model and chromatin immunoprecipitation assays, we find that both the cytosolic activation and nuclear transactivation phases of NF-B occur in LPS responsive THP-1 promonocytes with recruitment and binding of NF-B p65 at the IL-1 promoter. However, transcriptionally repressed LPS-tolerant THP-1 cells do not bind NF-B p65 at the IL-1 promoter, despite cytosolic activation and accumulation of p65 in the nucleus. In contrast, NF-B p50, which also accumulates in the nucleus, constitutively binds to the IL-1 promoter NF-B site in both LPS-responsive and LPS-tolerant cells. The level of p65 binding correlates with a binary shift in nucleosome remodeling between histone H3 phosphorylation at serine 10 and methylation of histone H3 at lysine 9. We conclude that LPS tolerance disrupts the transactivating stage of NF-B p65 and altered nucleosome remodeling at the IL-1 promoter in human leukocytes.
Introduction
Severe septicemia or septic shock (collectively referred to herein as septicemia) is a disease often caused by severe Gram-negative bacterial infection that is characterized by disseminated inflammation and coagulation with multiple organ failure (1, 2). Gram-negative bacterial cell wall LPS or endotoxin, as well as other microbial chemicals that act through the TLR, can initiate septicemia through its potent effects on the innate immune system (3). Blood leukocytes are crucial mediators of the innate immune response toward LPS and secrete numerous proinflammatory cytokines that include IL-1 and TNF-. Leukocyte hyperreactivity toward LPS can lead to the overexpression of these inflammatory cytokines that play a crucial role in inducing septicemia.
The initial burst of proinflammatory gene expression in response to LPS is short-lived and replaced by the prolonged expression of various anti-inflammatory genes. The continued down-regulation of proinflammatory mediators and continued production of anti-inflammatory proteins can lead to an immunosuppressed state that is characterized by LPS tolerance (4, 5, 6). These phenotypic changes in the innate immune response are often seen during animal and human septicemia, where the risk of secondary bacterial infection increases due to immune repression.
The NF-B family of transcription factors are major intracellular components for regulating immune and inflammatory responses, including regulating the expression of a variety of proinflammatory genes such as IL-1 (7). NF-B is the generic term used for transcription factors consisting of homo- or heterodimers of the Rel family of proteins. The Rel family consists of five proteins: p65 (RelA), RelB, cRel, p50, and p52 (8), with the active form of NF-B most commonly composed of p65 (RelA) and p50 heterodimers (9). p65 is the subunit responsible for the strongest transcriptional activating potential of NF-B (10), whereas the p50 subunit accounts for the strong DNA-binding affinity. It has also been shown that p50 homodimers can repress the gene expression of some NF-B-dependent genes (11). NF-B p65 and p50 modulate expression of >150 genes (12).
NF-B p65 and p50 play crucial roles in initiating transcription of these and many other genes involved in septicemia. Substantial data indicate that there are persistent elevations of nuclear p65 during the course of human septicemia in blood leukocytes and failing organs (13, 14, 15). Paradoxically, NF-B-dependent transcription of proinflammatory genes is persistently repressed during septicemia (16, 17). This suppression of innate immunity, which closely simulates classic LPS tolerance (16, 18), portends a poor clinical outcome (19).
The emergence of the chromatin immunoprecipitation assay (ChIP)3 stemmed from the revelation that histone proteins of chromatin associated with the packaging of DNA are also involved in the regulation of gene expression (reviewed in Ref. 20). The N termini of the four core histones (H2A, H2B, H3, H4), which are arranged in an octamer as a nucleosome, are susceptible to a variety of covalent modifications at a number of sites. The most common histone modification- associated transcriptional activation of immediate response genes include demethylation of Lys9, the phosphorylation of Ser10, and the acetylation of Lys9 and Lys14 on histone H3 (21, 22). The terms "binary switches" and "modification cassettes" have recently been used to describe a model of how histone modifications can regulate gene expression. In particular is the putative "phos/methyl switch," where phosphorylation of a histone residue adjacent to a methylation mark would cause a subsequent loss of binding of certain negative effectors, leading to the demethylation of that site and the recruitment of positive effector molecules that enhance gene expression (23). The ChIP assay can assess in vivo transcriptional events by observing effector protein binding and histone modifications to specific genes.
The objective of this study was to examine chromatin-associated molecular events that might repress NF-B-dependent gene expression in human leukocytes during septicemia. We investigated the in vivo binding patterns of p65 and p50 to NF-B sites in the promoter region of the IL-1 gene in LPS-responsive and LPS-tolerant human THP-1 promonocytes and the chromatin remodeling that accompanies this binding. The LPS tolerance observed in this model closely simulates the LPS refractory state observed in human septicemia monocytes (24) and polymorphonuclear neutrophils (16). We previously reported that LPS tolerance in the human promonocytic cell line THP-1 is associated with a decrease in LPS-induced IL-1 transcription (25). We further showed that, despite LPS-dependent NF-B translocation to the nucleus and DNA binding, as determined by in vitro EMSA, NF-B-dependent transcription is repressed in LPS-tolerant cells (26, 27).
In this report, we use immunoblots and ChIP assays to determine, in vivo, LPS-induced NF-B binding to the IL-1 promoter region and compare the binding of p65 and p50 in LPS-responsive and LPS-tolerant THP-1 cells. We show for the first time in LPS-tolerant cells that there is a specific disruption of the nuclear activation phase of p65 with decreased binding of this transcription factor to the IL-1 promoter. In contrast, p50 constitutively binds to the promoter of the IL-1 gene and no apparent LPS-induced changes in binding of p50 occurred in LPS-responsive and LPS-tolerant cells. We also found that histone H3 phosphorylation on Ser10 is significantly increased at the IL-1 promoter in normal cells treated with LPS compared with tolerant cells treated with LPS. In comparison, we show histone H3 methylation levels on Lys9 to be inversely correlated with NF-B p65 binding in LPS-stimulated normal cells, while maintaining steady methylated histone H3 levels (Lys9) in tolerant cells treated with LPS. Acetylation of histone H3 at Lys9 and Lys14, on the other hand, showed no significant changes in normal or LPS-tolerant cells treated with LPS. These results suggest that NF-B binding and the remodeling of promoter nucleosomes may play pivotal roles in regulating the expression of proinflammatory genes during septicemia and that these processes are disrupted during LPS tolerance.
Materials and Methods
Cell culture and stimulation
THP-1 cells obtained from the American Type Culture Collection were maintained in RPMI 1640 medium (Invitrogen Life Technologies) supplemented with 10 U/ml penicillin G, 10 μg/ml streptomycin, 2 mM L-glutamine, and 10% FBS (HyClone) at 37°C and 5% CO2 in a humidified incubator. LPS-tolerant THP-1 cells were made and are characterized in detail as described previously (25). Cells were rendered LPS tolerant by preincubation over 16 h with LPS (1.0 μg/ml Gram-negative LPS (Escherichia coli 0111:B4; Sigma-Aldrich). Normal and tolerant THP-1 cells (1.0 x 106 cells/condition) were washed once with FBS-free RPMI 1640 medium, then resuspended in FBS supplemented RPMI 1640 medium at 1 x 106 cells/ml and stimulated with LPS (1.0 μg/ml) for the indicated times. Low passage number and log-phase cells were used for all experiments.
ChIP assay
To assess p65 and p50 binding and histone modifications to the IL-1 promoter, ChIP assays (Upstate Biotechnology) were performed according to the manufacturer’s instructions with the following modifications. LPS-stimulated cells were fixed by adding formaldehyde (HCHO, from a 37% HCHO/10% methanol stock; Calbiochem) into the medium for a final formaldehyde concentration of 1% and incubated at room temperature for 10 min. with gentle shaking. Incubation with the lysis buffer was extended to 25 min at 4°C. The chromatin was sheared by sonication using a Branson 250 sonicator with microtip at a power setting of 2 and 40% duty cycle. We found that 2 x 5-s bursts at this setting generated fragments of 0.5–1.5 kb. The samples were placed on ice for 1 min between sonication bursts. Each condition was divided into two samples, providing a preimmunoprecipitation or "input" sample that was not incubated with specific Abs, and an immunoprecipitated sample that was incubated overnight with Abs specific for either p50, p65, phospho-histone H3 (Ser10), total histone H3, acetylated histone H3 (Lys9, Lys14) (Santa Cruz Biotechnology), and methyl-histone H3 (Lys9) (Upstate Biotechnology). Abs that recognized different p50 or p65 epitopes were used with similar results. Preimmune IgG was used as a negative control (data not shown). DNA was isolated by phenol/chloroform extraction, ethanol precipitated, and resuspended in 20 μl of dH2O. DNA obtained by the ChIP procedure was first subjected to semiquantitative PCR with the products visualized on agarose gels. Four microliters of immunoprecipitated DNA was used for each PCR. The following primers specific for the IL-1 promoter were used: promoter 5' primer, 5'-cactcttccactccctcc-3'; promoter 3' primer, 5'-agcctcaaacccttcctc-3'; nonspecific 5' primer, 5'-aaaggaggtgtggaaccag-3'; nonspecific 3' primer, 5'-cgagcaatgtaaaatgaggaag-3' (Integrated DNA Technologies). Their coordinates (Table I) and positions relative to the IL-1 NF-B binding and transcription start sites are shown in Fig. 1. PCR products were quantified using a Typhoon 8600 Scanner (Molecular Dynamics and Amersham Pharmaceutical Biotech), normalized to the input DNA and expressed as an average fold increase in binding relative to unstimulated samples. PCR was also performed on samples containing the IL-1 promoter DNA of known concentrations and show that our PCR conditions were within the linear range (data not shown). Statistical analysis (t test) and graphic presentations from the results of at least three ChIP assays were performed using Microsoft Excel XP.
Quantitative real-time PCR
For more precise quantification of the ChIP results, real-time PCR was used. The ChIPed DNA for each treatment was analyzed quantitatively for the amplification of the IL-1 promoter using specific primers (forward, 5'-cgtgggaaaatccagtattttaatg-3' and reverse, 5'-caaatgtatcaccatgcaaatatgc-3') and probe (5'-6-FAM-acatcaactgcacaacgattgtcaggaaaa-TAMRA-3') designed by Primer Express R 1.0 software (Applied Biosystems). Samples from at least three independent immunoprecipitations were analyzed by quantitative PCR. PCR products were continuously measured by means of an ABI Prism 7000 during 40 cycles. All data were normalized to the input DNA and DNA from untreated THP-1 cells.
Western blot analysis
Nuclear extracts were used to determine the nuclear levels of NF-B p65 and p50 using the protocol described previously (28) with the following modifications. Fifty-microgram proteins of nuclear extract protein were separated by SDS-PAGE and transferred to an Immuno-Blot polyvinylidene difluoride membrane (Bio-Rad). NF-B p65 or p50 Abs (Santa Cruz Biotechnology) were used to visualize and quantify protein levels using ImageQuant software (Molecular Dynamics and Amersham Pharmaceutical Biotech). Statistical analysis (t test) and graphic presentations were performed using Microsoft Excel XP statistical analysis software.
Results
NF-B p65 binding to the IL-1 promoter is induced by LPS in normal cells and repressed in LPS-tolerant cells
NF-B p65, the subunit responsible for the strong transcription-activating potential of NF-B, has been shown by EMSA to bind rapidly to the NF-B sites on IL-1 when activated by LPS (7, 26). We also demonstrated that NF-B p65 in tolerant cells retains LPS-dependent DNA binding as determined by EMSA (26). In this study, we used the ChIP assay to examine the in vivo binding of p65 to NF-B sites on the promoter region of the IL-1 gene. Fig. 1 shows the NF-B binding site relative to the start of the IL-1 gene. As shown in Fig. 2, p65 is rapidly and transiently recruited to the promoter NF-B sites in response to LPS stimulation with an almost 10-fold increase in binding within 1 h (Fig. 2A, compare Normal 0 and Normal 0.5). In contrast, LPS tolerant cells showed significantly reduced p65 binding to the promoter after LPS treatment (Fig. 2A, compare Normal 0.5 and Tolerant 0.5). We further examined and quantified p65 binding to the IL-1 promoter region using real-time PCR (Fig. 2B) with similar results. We also show, as a negative control, that a region of the IL-1 gene that does not contain any NF-B binding sites does not bind p65 with LPS stimulation (Fig. 2C). Finally, Western blot analyses of nuclear extracts show that upon LPS stimulation both LPS-responsive and LPS-tolerant phenotypes exhibit an increase in nuclear levels on NF-B p65. Furthermore, nuclear levels of NF-B p65 are not significantly reduced in the LPS-tolerant phenotype (Fig. 2D). These results demonstrate that despite the ability of p65 to translocate into the nucleus in LPS-treated tolerant cells, its ability to bind the IL-1 promoter is decreased in tolerance. This suggests that the inability of p65 to bind regulatory regions in the IL-1 gene may be responsible, in part, for the decreased expression of IL-1 observed in tolerant cells. The contrast between these in vivo results and our previous findings (26) also demonstrate the limitations of EMSA in investigating the interaction between transcription factors and regulatory regions of specific gene targets.
p50 binding to the IL-1 promoter is constitutive and shows no apparent change with LPS stimulation in normal and tolerant cells
Using EMSA, we had previously shown that DNA binding of p50 is significantly increased in response to LPS in both normal and tolerant cells. To extend these studies, we performed a ChIP analysis with the same conditions described for p65 to examine the in vivo binding patterns of p50 to the promoter region of the IL-1 gene. Fig. 3A, shows p50 binding to the NF-B site in the promoter and the results indicate that p50 binds to the IL-1 promoter constitutively. Upon stimulation with LPS, no significant change in p50 binding to the IL-1 promoter was observed in either normal or LPS-tolerant cells. Again, we further analyzed the p50 binding results using real-time PCR (Fig. 3B) with similar results. We also show that a region of the IL-1 gene that does not contain any NF-B binding sites does not bind p50 with LPS stimulation (Fig. 3C). Western blot analysis of nuclear levels of NF-B p50 demonstrates relatively low amounts of nuclear p50 in untreated cells; however, p50 is rapidly translocated into the nucleus soon after LPS stimulation in normal cells. In LPS-tolerant cells, nuclear p50 levels appear to be constitutively present and are further increased upon additional stimulation with LPS (Fig. 3D). These results demonstrate that LPS-tolerant cells not only retain but also have an enhanced ability to translocate NF-B p50 into the nucleus. These elevated levels of p50 are not associated with detectable increases in p50 binding to DNA in tolerant cells, as assessed by ChIP; however, ChIP would not detect exchange of homodimers of p50 to heterodimers of p65 and a return of homodimers of p50 in the LPS-tolerant phenotype. As an additional control, preimmune, nonrelevant Abs were used in ChIP assays and demonstrated the specificity of the immune complexes precipitated with the NF-B Abs (data not shown).
A binary shift in nucleosome remodeling at the IL-1 promoter is present between histone H3 phosphorylation and histone H3 methylation during LPS stimulation of normal cells, but absent in LPS-tolerant cells; however, no significant changes in H3 acetylation were observed
To this point, we have observed a decrease in NF-B p65 binding in LPS-tolerant cells despite the increased presence of p65 in the nucleus. To further explore the mechanisms that would regulate gene expression and p65 binding, we looked at chromatin remodeling at the IL-1 promoter, since it has been well established that histone modifications play an important role in regulating transcriptional activity. We used ChIP assays to analyze the phosphorylation (Ser10), methylation (Lys9), and acetylation (Lys9, Lys14) of histone H3 at the IL-1 promoter in LPS-responsive and LPS-tolerant cells. We found that in normal untreated cells, histone H3 at Ser10 was constitutively phosphorylated at the promoter and upon LPS stimulation observed a sharp increase in H3 phosphorylation that peaked at 1 h after LPS stimulation. In contrast, the IL-1 promoter showed significantly reduced histone phosphorylation H3 (Ser10) in tolerant cells at corresponding times after LPS. These values were compared with the "Total H3" values obtained by ChIP analysis for total histone H3 (Fig. 4A) We performed further ChIP analysis directed against diacetylated histone H3 (Lys9, Lys14), specifically on the IL-1 promoter. We found that histone H3 around the IL-1 promoter region were constitutively acetylated and did not change significantly in normal or tolerant cells treated with LPS (Fig. 4C). These values were compared with the Total H3 values obtained by ChIP analysis for total histone H3 (Fig. 4C). When we examined the methylation levels on Lys9 of histone H3 around the IL-1 promoter, we found that methylated histone H3 levels Lys9 were highest in normal untreated THP-1 cells. With LPS stimulation, we observed a significant decrease in histone H3 methylation that reached its lowest peak at 1 h. Remethylation of histone H3 on Lys9 gradually appears by 3 h of LPS stimulation, but does not appear to return back to a basal methylation state. In LPS-tolerant cells however, histone H3 methylation on Lys9 remains steady and relatively unchanged around the IL-1 promoter throughout LPS treatment (Fig. 4F). These results exhibit an inverse relationship of H3 methylation to both histone H3 phosphorylation (Ser10) and NF-B p65 binding. These values were also compared with the Total H3 values obtained by ChIP analysis for total histone H3 (Fig. 4F, upper panel). ChIPed DNA for the phosphorylated histone H3, acetylated histone H3, "total" H3, and methylated histone H3 were also quantitated using real-time PCR (Fig. 4, B, D, E, and G, respectively) with similar results.
Taken together, these results demonstrate that there is nuclear translocation of both p65 and p50 in response to LPS in both normal and tolerant cells, and that binding of p65 to the IL-1 promoter increases in LPS-responsive but not LPS-tolerant cells. p50 binding to the promoter is detected in the constitutive unstimulated state as well as in LPS-responsive and LPS-tolerant cells. Thus, p50 binding appears to remain unchanged throughout LPS treatment despite LPS-induced increases in nuclear p50 in tolerant cells (Fig. 3D). These in vivo results contrast our previous results using EMSA (26). We also found that LPS induced phosphorylation of histone H3 on Ser10 and coincident demethylation of histone H3 on Lys9 at the transcriptionally active IL-1 promoter. These modifications were not observed at the inactive IL-1 promoter in tolerant cells.
Discussion
NF-B regulation of transcription by the classical or canonical pathway can be divided into two phases (10, 29). First is a cytosolic phase in which there is IB kinase -dependent activation of IB followed by ubiquitin-directed IB degradation that allows translocation of p50 and p65 into the nucleus. The second or nuclear phase of NF-B regulation involves events that result in derepression NF-B-dependent promoters, chromatin remodeling, and binding of transcriptionally active p65 and p50 as a heterodimer to cognate DNA on enhancer and promoter elements. Assembly of the full promoter transcriptome follows this with activation of RNA polymerase II (30). This canonical mechanism for NF-B activation is fundamental in transactivating genes expressed by innate immune cells, and these gene products are critically involved in host protection and host injury, including septicemia (31). However, soon after septicemia is initiated, the NF-B pathway is disrupted at the level of transcription (4, 6, 16), and this occurs coincident with repression of proinflammatory genes such as TNF- and IL-1. Thus, these and other proinflammatory genes are reprogrammed during septicemia to be LPS tolerant. Surprisingly, there is clear evidence in circulating leukocytes and tissues showing chronic activation of NF-B with nuclear accumulation of p65 during sepsis (13, 15, 32), and these studies often use the inclusive and somewhat misleading term "activated NF-B."
To our knowledge we report the first observation that may explain the activation of classical pathway of cytosolic NF-B with accumulation of p65 in the nucleus, while there is repression of the nuclear phase of NF-B-dependent transcription. Using IL-1 as the NF-B-dependent gene and THP-1 cells as a model of septicemia leukocytes with LPS tolerance (the IL-1 gene is consistently repressed in human blood sepsis leukocytes, but variable in murine leukocytes), we show that the IL-1 promoter is disrupted, both at the level of p65 binding and nucleosome remodeling during LPS tolerance. LPS tolerance with repression of IL-1 transcription is a consistent feature of septicemia leukocytes studied ex vivo (4, 18).
We also provide the novel observation that p50 is bound to the IL-1 promoter in the basal, LPS-responsive, and LPS-tolerant states, without any observable changes. Our results support recent findings suggesting that p50 has evolved to constitutively bind to promoters of inflammatory genes without disturbing the nucleosomes or requiring their remodeling (33), thus providing rapid access to other transcription factors like p65. With pronounced increases in nuclear p50 levels in LPS-tolerant cells and the lack of any observable change in p50 binding to the IL-1 promoter, it is likely that the bulk of the p50 binding observed in LPS-tolerant cells occurs as a p50 homodimer, whereas the majority of the p50 found in LPS-responsive cells occurs as part of a p65-p50 heterodimer complex.
NF-B-dependent genes are usually controlled by p50 or p52 homodimers that repress transcription that depends on corepressors that bind to the homodimers (discussed in Ref. 34). It has been clearly shown that the p50 plays a role in LPS tolerance and that p50 homodimers predominate in the LPS-tolerant phenotype (14, 35, 36). One potential explanation for LPS tolerance at the level of the gene is formation of a repressor complex, which includes inhibitor corepressors and monomers or homodimers of p50, that are reassembled after the initial LPS response of septicemia stress (37, 38). We are currently exploring the possibility that other NF-B proteins and/or other interacting transcription factors may be assembled at the IL-1 promoter in LPS-tolerant cells.
Another level of regulation that may participate in transcription repression in LPS tolerance is chromatin remodeling of the promoter by histone modification (39). Our data show increased phosphorylation of Ser10 on histone H3 in the LPS-responsive phenotype and a hypophosphorylated state of this amino acid in the LPS-tolerant cell. Phosphorylation of H3 at Ser10 is a critical event in nucleosome remodeling, and our data further show that phosphorylation of H3 on Ser10 occurs in conjunction with demethylation of H3 on Lys9 (40). This methylation/phosphorylation shift supports the model that was recently offered as the hypothetical binary shift paradigm for chromatin remodeling (23). Although we did not observe demonstrable changes in acetylation of Lys9 or Lys14, it is possible that changes in acetylation occur at other residues or on other histone tails, or that histone acetylation is not required for regulating transcription factor binding to inflammatory genes. It is unclear whether activation of specific demethylases is a prerequisite to phosphorylation of H3 Ser10 or whether H3 Ser10 phosphorylation results in demethylation. Although nucleosome remodeling plays an important role in LPS tolerance at the IL-1 promoter, it is unlikely to be the only mechanism responsible for tolerance, since LPS tolerance disrupts transcription of both transiently and stably transfected reporter constructs (26). The higher order structure of transient transfected genes is distinct from the native chromatin organization (41, 42).
Of further potential importance are more intracellular signaling components such as kinases (43, 44, 45, 46) or phosphatases (47). p38 MAPK (43, 48), probably acting via another kinase such as mitogen-stimulated kinase and IB kinase (49, 50) are implicated in both phosphorylating p65 on Ser536 for enhanced transactivation, as well as phosphorylating histone H3 on Ser10 for nucleosome remodeling at or near the promoters of NF-B-dependent genes. Disruption of p38 kinase activity has been implicated in in vitro and in vivo models of LPS tolerance (reviewed in5), and others and we (data not shown) have found p38 inactive in LPS-tolerant THP-1 cells.
Clearly, the generation of an LPS-tolerant phenotype is complex and multifaceted. This report focuses on nuclear events at or near the promoter nucleosome of inflammatory genes. There is reprogramming of proximal cytosolic mediators such as IRAK monomyeloic, MyD88, Tollip, SHIP, single Ig IL-1R related protein, and perhaps suppressor of cytokine signaling 1 and 3 during LPS tolerance (51, 52, 53). The complexity of the LPS-tolerant phenotypes likely extends from the cell membrane into the nucleus, where the final pathways for repressing proinflammatory genes and sustaining the expression of anti-inflammatory genes operate. Fig. 5 is a schematic model of suppression of nuclear events involving NF-B at the IL-1 promoter of LPS-responsive vs LPS-tolerant THP-1 cells.
In summary, NF-B p65 binding and chromatin remodeling at the IL-1 promoter gene are differentially altered in the LPS-tolerant, when compared with the LPS-responsive, THP-1 monocytes. NF-B is activated in the cytosol of both LPS-responsive and LPS-tolerant cells with translocation of p65 and p50 into the nucleus. NF-B p65 binding to the IL-1 promoter is rapidly and transiently increased in LPS-responsive monocytes, whereas its binding to the IL-1 promoter is significantly limited in LPS-tolerant cells, an observation that correlates with repressed transcription of IL-1. In contrast, NF-B p50 binding is constitutive and is apparently unchanged by LPS stimulation in normal or tolerant cells. In LPS-responsive cells, there is also a binary shift in nucleosome remodeling with histone H3 demethylation of Lys9 and increased histone H3 phosphorylation of Ser10, when significant NF-B p65 binding is observed. However, in LPS-tolerant cells where NF-B p65 binding levels are reduced, there are concomitant reductions in phosphorylated histone H3 on Ser10 and steady levels of methylated histone H3 at Lys9 are retained around the IL-1 promoter. These results suggest that promoter-specific alterations contribute to the LPS-tolerant phenotype and may contribute to the suppressed innate immune response in septicemia. Analysis of promoter regulatory events of distinct genes during LPS tolerance might provide more targeted approaches to altering the disastrous outcome in septicemia.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grant AI-09169 (to C.E.M.) and American Lung Association Grants RG-035-N and RG-065-NL and American Heart Association Grant 0051211U (to B.K.Y.), and National Institutes of Allergy and Infectious Diseases AI-50089 (to L.L.).
2 Address correspondence and reprint requests to Dr. Barbara K. Yoza, Section of Molecular Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27157. E-mail address: byoza@wfubmc.edu
3 Abbreviation used in this paper: ChIP, chromatin immunoprecipitation.
Received for publication January 24, 2005. Accepted for publication April 12, 2005.
References
Angus, D. C., R. S. Wax. 2001. Epidemiology of sepsis: an update. Crit. Care Med. 29: S109-S116.
Bone, R. C., C. J. Grodzin, R. A. Balk. 1997. Sepsis: a new hypothesis for pathogenesis of the disease process. Chest 112: 235-243.
Beutler, B., K. Hoebe, X. Du, R. J. Ulevitch. 2003. How we detect microbes and respond to them: the Toll-like receptors and their transducers. J. Leukocyte Biol. 74: 479-485.
Cavaillon, J. M., C. Fitting, M. dib-Conquy. 2004. Mechanisms of immunodysregulation in sepsis. Contrib. Nephrol. 144: 76-93.
Fan, H., J. A. Cook. 2004. Molecular mechanisms of endotoxin tolerance. J. Endotoxin Res. 10: 71-84.
West, M. A., W. Heagy. 2002. Endotoxin tolerance: a review. Crit. Care Med. 30: S64-S73.
Cogswell, J. P., M. M. Godlevski, G. B. Wisely, W. C. Clay, L. M. Leesnitzer, J. P. Ways, J. G. Gray. 1994. NF-B regulates IL-1 transcription through a consensus NF-B binding site and a nonconsensus CRE-like site. J. Immunol. 153: 712-723.
Baldwin, A. S.. 2004. NFB: biology and pathology–Keystone Symposium. 11–16 January 2004, Snowbird, UT, USA. IDrugs 7: 215-216.
Baldwin, A. S.. 2001. Control of oncogenesis and cancer therapy resistance by the transcription factor NF-B. J. Clin. Invest. 107: 241-246.(Christopher Chan*, Liwu L)
The NF-B family plays a crucial role in the pathogenesis of highly lethal septicemia by modulating transcription of many innate and adaptive immunity genes. Two phases of NF-B activation occur: cytosolic activation and nuclear transactivation. Septicemia with multiorgan failure is associated with chronic activation of cytosolic NF-B with translocation and accumulation of increased levels of nuclear p65 in blood leukocytes. Paradoxically, NF-B-dependent transcription of many proinflammatory genes responding to bacterial LPS endotoxin (LPS) is persistently repressed during septicemia; this phenomenon of LPS tolerance is associated with immunosuppression and poor prognosis. This report suggests an explanation for this paradox. Using an in vitro human leukocyte model and chromatin immunoprecipitation assays, we find that both the cytosolic activation and nuclear transactivation phases of NF-B occur in LPS responsive THP-1 promonocytes with recruitment and binding of NF-B p65 at the IL-1 promoter. However, transcriptionally repressed LPS-tolerant THP-1 cells do not bind NF-B p65 at the IL-1 promoter, despite cytosolic activation and accumulation of p65 in the nucleus. In contrast, NF-B p50, which also accumulates in the nucleus, constitutively binds to the IL-1 promoter NF-B site in both LPS-responsive and LPS-tolerant cells. The level of p65 binding correlates with a binary shift in nucleosome remodeling between histone H3 phosphorylation at serine 10 and methylation of histone H3 at lysine 9. We conclude that LPS tolerance disrupts the transactivating stage of NF-B p65 and altered nucleosome remodeling at the IL-1 promoter in human leukocytes.
Introduction
Severe septicemia or septic shock (collectively referred to herein as septicemia) is a disease often caused by severe Gram-negative bacterial infection that is characterized by disseminated inflammation and coagulation with multiple organ failure (1, 2). Gram-negative bacterial cell wall LPS or endotoxin, as well as other microbial chemicals that act through the TLR, can initiate septicemia through its potent effects on the innate immune system (3). Blood leukocytes are crucial mediators of the innate immune response toward LPS and secrete numerous proinflammatory cytokines that include IL-1 and TNF-. Leukocyte hyperreactivity toward LPS can lead to the overexpression of these inflammatory cytokines that play a crucial role in inducing septicemia.
The initial burst of proinflammatory gene expression in response to LPS is short-lived and replaced by the prolonged expression of various anti-inflammatory genes. The continued down-regulation of proinflammatory mediators and continued production of anti-inflammatory proteins can lead to an immunosuppressed state that is characterized by LPS tolerance (4, 5, 6). These phenotypic changes in the innate immune response are often seen during animal and human septicemia, where the risk of secondary bacterial infection increases due to immune repression.
The NF-B family of transcription factors are major intracellular components for regulating immune and inflammatory responses, including regulating the expression of a variety of proinflammatory genes such as IL-1 (7). NF-B is the generic term used for transcription factors consisting of homo- or heterodimers of the Rel family of proteins. The Rel family consists of five proteins: p65 (RelA), RelB, cRel, p50, and p52 (8), with the active form of NF-B most commonly composed of p65 (RelA) and p50 heterodimers (9). p65 is the subunit responsible for the strongest transcriptional activating potential of NF-B (10), whereas the p50 subunit accounts for the strong DNA-binding affinity. It has also been shown that p50 homodimers can repress the gene expression of some NF-B-dependent genes (11). NF-B p65 and p50 modulate expression of >150 genes (12).
NF-B p65 and p50 play crucial roles in initiating transcription of these and many other genes involved in septicemia. Substantial data indicate that there are persistent elevations of nuclear p65 during the course of human septicemia in blood leukocytes and failing organs (13, 14, 15). Paradoxically, NF-B-dependent transcription of proinflammatory genes is persistently repressed during septicemia (16, 17). This suppression of innate immunity, which closely simulates classic LPS tolerance (16, 18), portends a poor clinical outcome (19).
The emergence of the chromatin immunoprecipitation assay (ChIP)3 stemmed from the revelation that histone proteins of chromatin associated with the packaging of DNA are also involved in the regulation of gene expression (reviewed in Ref. 20). The N termini of the four core histones (H2A, H2B, H3, H4), which are arranged in an octamer as a nucleosome, are susceptible to a variety of covalent modifications at a number of sites. The most common histone modification- associated transcriptional activation of immediate response genes include demethylation of Lys9, the phosphorylation of Ser10, and the acetylation of Lys9 and Lys14 on histone H3 (21, 22). The terms "binary switches" and "modification cassettes" have recently been used to describe a model of how histone modifications can regulate gene expression. In particular is the putative "phos/methyl switch," where phosphorylation of a histone residue adjacent to a methylation mark would cause a subsequent loss of binding of certain negative effectors, leading to the demethylation of that site and the recruitment of positive effector molecules that enhance gene expression (23). The ChIP assay can assess in vivo transcriptional events by observing effector protein binding and histone modifications to specific genes.
The objective of this study was to examine chromatin-associated molecular events that might repress NF-B-dependent gene expression in human leukocytes during septicemia. We investigated the in vivo binding patterns of p65 and p50 to NF-B sites in the promoter region of the IL-1 gene in LPS-responsive and LPS-tolerant human THP-1 promonocytes and the chromatin remodeling that accompanies this binding. The LPS tolerance observed in this model closely simulates the LPS refractory state observed in human septicemia monocytes (24) and polymorphonuclear neutrophils (16). We previously reported that LPS tolerance in the human promonocytic cell line THP-1 is associated with a decrease in LPS-induced IL-1 transcription (25). We further showed that, despite LPS-dependent NF-B translocation to the nucleus and DNA binding, as determined by in vitro EMSA, NF-B-dependent transcription is repressed in LPS-tolerant cells (26, 27).
In this report, we use immunoblots and ChIP assays to determine, in vivo, LPS-induced NF-B binding to the IL-1 promoter region and compare the binding of p65 and p50 in LPS-responsive and LPS-tolerant THP-1 cells. We show for the first time in LPS-tolerant cells that there is a specific disruption of the nuclear activation phase of p65 with decreased binding of this transcription factor to the IL-1 promoter. In contrast, p50 constitutively binds to the promoter of the IL-1 gene and no apparent LPS-induced changes in binding of p50 occurred in LPS-responsive and LPS-tolerant cells. We also found that histone H3 phosphorylation on Ser10 is significantly increased at the IL-1 promoter in normal cells treated with LPS compared with tolerant cells treated with LPS. In comparison, we show histone H3 methylation levels on Lys9 to be inversely correlated with NF-B p65 binding in LPS-stimulated normal cells, while maintaining steady methylated histone H3 levels (Lys9) in tolerant cells treated with LPS. Acetylation of histone H3 at Lys9 and Lys14, on the other hand, showed no significant changes in normal or LPS-tolerant cells treated with LPS. These results suggest that NF-B binding and the remodeling of promoter nucleosomes may play pivotal roles in regulating the expression of proinflammatory genes during septicemia and that these processes are disrupted during LPS tolerance.
Materials and Methods
Cell culture and stimulation
THP-1 cells obtained from the American Type Culture Collection were maintained in RPMI 1640 medium (Invitrogen Life Technologies) supplemented with 10 U/ml penicillin G, 10 μg/ml streptomycin, 2 mM L-glutamine, and 10% FBS (HyClone) at 37°C and 5% CO2 in a humidified incubator. LPS-tolerant THP-1 cells were made and are characterized in detail as described previously (25). Cells were rendered LPS tolerant by preincubation over 16 h with LPS (1.0 μg/ml Gram-negative LPS (Escherichia coli 0111:B4; Sigma-Aldrich). Normal and tolerant THP-1 cells (1.0 x 106 cells/condition) were washed once with FBS-free RPMI 1640 medium, then resuspended in FBS supplemented RPMI 1640 medium at 1 x 106 cells/ml and stimulated with LPS (1.0 μg/ml) for the indicated times. Low passage number and log-phase cells were used for all experiments.
ChIP assay
To assess p65 and p50 binding and histone modifications to the IL-1 promoter, ChIP assays (Upstate Biotechnology) were performed according to the manufacturer’s instructions with the following modifications. LPS-stimulated cells were fixed by adding formaldehyde (HCHO, from a 37% HCHO/10% methanol stock; Calbiochem) into the medium for a final formaldehyde concentration of 1% and incubated at room temperature for 10 min. with gentle shaking. Incubation with the lysis buffer was extended to 25 min at 4°C. The chromatin was sheared by sonication using a Branson 250 sonicator with microtip at a power setting of 2 and 40% duty cycle. We found that 2 x 5-s bursts at this setting generated fragments of 0.5–1.5 kb. The samples were placed on ice for 1 min between sonication bursts. Each condition was divided into two samples, providing a preimmunoprecipitation or "input" sample that was not incubated with specific Abs, and an immunoprecipitated sample that was incubated overnight with Abs specific for either p50, p65, phospho-histone H3 (Ser10), total histone H3, acetylated histone H3 (Lys9, Lys14) (Santa Cruz Biotechnology), and methyl-histone H3 (Lys9) (Upstate Biotechnology). Abs that recognized different p50 or p65 epitopes were used with similar results. Preimmune IgG was used as a negative control (data not shown). DNA was isolated by phenol/chloroform extraction, ethanol precipitated, and resuspended in 20 μl of dH2O. DNA obtained by the ChIP procedure was first subjected to semiquantitative PCR with the products visualized on agarose gels. Four microliters of immunoprecipitated DNA was used for each PCR. The following primers specific for the IL-1 promoter were used: promoter 5' primer, 5'-cactcttccactccctcc-3'; promoter 3' primer, 5'-agcctcaaacccttcctc-3'; nonspecific 5' primer, 5'-aaaggaggtgtggaaccag-3'; nonspecific 3' primer, 5'-cgagcaatgtaaaatgaggaag-3' (Integrated DNA Technologies). Their coordinates (Table I) and positions relative to the IL-1 NF-B binding and transcription start sites are shown in Fig. 1. PCR products were quantified using a Typhoon 8600 Scanner (Molecular Dynamics and Amersham Pharmaceutical Biotech), normalized to the input DNA and expressed as an average fold increase in binding relative to unstimulated samples. PCR was also performed on samples containing the IL-1 promoter DNA of known concentrations and show that our PCR conditions were within the linear range (data not shown). Statistical analysis (t test) and graphic presentations from the results of at least three ChIP assays were performed using Microsoft Excel XP.
Quantitative real-time PCR
For more precise quantification of the ChIP results, real-time PCR was used. The ChIPed DNA for each treatment was analyzed quantitatively for the amplification of the IL-1 promoter using specific primers (forward, 5'-cgtgggaaaatccagtattttaatg-3' and reverse, 5'-caaatgtatcaccatgcaaatatgc-3') and probe (5'-6-FAM-acatcaactgcacaacgattgtcaggaaaa-TAMRA-3') designed by Primer Express R 1.0 software (Applied Biosystems). Samples from at least three independent immunoprecipitations were analyzed by quantitative PCR. PCR products were continuously measured by means of an ABI Prism 7000 during 40 cycles. All data were normalized to the input DNA and DNA from untreated THP-1 cells.
Western blot analysis
Nuclear extracts were used to determine the nuclear levels of NF-B p65 and p50 using the protocol described previously (28) with the following modifications. Fifty-microgram proteins of nuclear extract protein were separated by SDS-PAGE and transferred to an Immuno-Blot polyvinylidene difluoride membrane (Bio-Rad). NF-B p65 or p50 Abs (Santa Cruz Biotechnology) were used to visualize and quantify protein levels using ImageQuant software (Molecular Dynamics and Amersham Pharmaceutical Biotech). Statistical analysis (t test) and graphic presentations were performed using Microsoft Excel XP statistical analysis software.
Results
NF-B p65 binding to the IL-1 promoter is induced by LPS in normal cells and repressed in LPS-tolerant cells
NF-B p65, the subunit responsible for the strong transcription-activating potential of NF-B, has been shown by EMSA to bind rapidly to the NF-B sites on IL-1 when activated by LPS (7, 26). We also demonstrated that NF-B p65 in tolerant cells retains LPS-dependent DNA binding as determined by EMSA (26). In this study, we used the ChIP assay to examine the in vivo binding of p65 to NF-B sites on the promoter region of the IL-1 gene. Fig. 1 shows the NF-B binding site relative to the start of the IL-1 gene. As shown in Fig. 2, p65 is rapidly and transiently recruited to the promoter NF-B sites in response to LPS stimulation with an almost 10-fold increase in binding within 1 h (Fig. 2A, compare Normal 0 and Normal 0.5). In contrast, LPS tolerant cells showed significantly reduced p65 binding to the promoter after LPS treatment (Fig. 2A, compare Normal 0.5 and Tolerant 0.5). We further examined and quantified p65 binding to the IL-1 promoter region using real-time PCR (Fig. 2B) with similar results. We also show, as a negative control, that a region of the IL-1 gene that does not contain any NF-B binding sites does not bind p65 with LPS stimulation (Fig. 2C). Finally, Western blot analyses of nuclear extracts show that upon LPS stimulation both LPS-responsive and LPS-tolerant phenotypes exhibit an increase in nuclear levels on NF-B p65. Furthermore, nuclear levels of NF-B p65 are not significantly reduced in the LPS-tolerant phenotype (Fig. 2D). These results demonstrate that despite the ability of p65 to translocate into the nucleus in LPS-treated tolerant cells, its ability to bind the IL-1 promoter is decreased in tolerance. This suggests that the inability of p65 to bind regulatory regions in the IL-1 gene may be responsible, in part, for the decreased expression of IL-1 observed in tolerant cells. The contrast between these in vivo results and our previous findings (26) also demonstrate the limitations of EMSA in investigating the interaction between transcription factors and regulatory regions of specific gene targets.
p50 binding to the IL-1 promoter is constitutive and shows no apparent change with LPS stimulation in normal and tolerant cells
Using EMSA, we had previously shown that DNA binding of p50 is significantly increased in response to LPS in both normal and tolerant cells. To extend these studies, we performed a ChIP analysis with the same conditions described for p65 to examine the in vivo binding patterns of p50 to the promoter region of the IL-1 gene. Fig. 3A, shows p50 binding to the NF-B site in the promoter and the results indicate that p50 binds to the IL-1 promoter constitutively. Upon stimulation with LPS, no significant change in p50 binding to the IL-1 promoter was observed in either normal or LPS-tolerant cells. Again, we further analyzed the p50 binding results using real-time PCR (Fig. 3B) with similar results. We also show that a region of the IL-1 gene that does not contain any NF-B binding sites does not bind p50 with LPS stimulation (Fig. 3C). Western blot analysis of nuclear levels of NF-B p50 demonstrates relatively low amounts of nuclear p50 in untreated cells; however, p50 is rapidly translocated into the nucleus soon after LPS stimulation in normal cells. In LPS-tolerant cells, nuclear p50 levels appear to be constitutively present and are further increased upon additional stimulation with LPS (Fig. 3D). These results demonstrate that LPS-tolerant cells not only retain but also have an enhanced ability to translocate NF-B p50 into the nucleus. These elevated levels of p50 are not associated with detectable increases in p50 binding to DNA in tolerant cells, as assessed by ChIP; however, ChIP would not detect exchange of homodimers of p50 to heterodimers of p65 and a return of homodimers of p50 in the LPS-tolerant phenotype. As an additional control, preimmune, nonrelevant Abs were used in ChIP assays and demonstrated the specificity of the immune complexes precipitated with the NF-B Abs (data not shown).
A binary shift in nucleosome remodeling at the IL-1 promoter is present between histone H3 phosphorylation and histone H3 methylation during LPS stimulation of normal cells, but absent in LPS-tolerant cells; however, no significant changes in H3 acetylation were observed
To this point, we have observed a decrease in NF-B p65 binding in LPS-tolerant cells despite the increased presence of p65 in the nucleus. To further explore the mechanisms that would regulate gene expression and p65 binding, we looked at chromatin remodeling at the IL-1 promoter, since it has been well established that histone modifications play an important role in regulating transcriptional activity. We used ChIP assays to analyze the phosphorylation (Ser10), methylation (Lys9), and acetylation (Lys9, Lys14) of histone H3 at the IL-1 promoter in LPS-responsive and LPS-tolerant cells. We found that in normal untreated cells, histone H3 at Ser10 was constitutively phosphorylated at the promoter and upon LPS stimulation observed a sharp increase in H3 phosphorylation that peaked at 1 h after LPS stimulation. In contrast, the IL-1 promoter showed significantly reduced histone phosphorylation H3 (Ser10) in tolerant cells at corresponding times after LPS. These values were compared with the "Total H3" values obtained by ChIP analysis for total histone H3 (Fig. 4A) We performed further ChIP analysis directed against diacetylated histone H3 (Lys9, Lys14), specifically on the IL-1 promoter. We found that histone H3 around the IL-1 promoter region were constitutively acetylated and did not change significantly in normal or tolerant cells treated with LPS (Fig. 4C). These values were compared with the Total H3 values obtained by ChIP analysis for total histone H3 (Fig. 4C). When we examined the methylation levels on Lys9 of histone H3 around the IL-1 promoter, we found that methylated histone H3 levels Lys9 were highest in normal untreated THP-1 cells. With LPS stimulation, we observed a significant decrease in histone H3 methylation that reached its lowest peak at 1 h. Remethylation of histone H3 on Lys9 gradually appears by 3 h of LPS stimulation, but does not appear to return back to a basal methylation state. In LPS-tolerant cells however, histone H3 methylation on Lys9 remains steady and relatively unchanged around the IL-1 promoter throughout LPS treatment (Fig. 4F). These results exhibit an inverse relationship of H3 methylation to both histone H3 phosphorylation (Ser10) and NF-B p65 binding. These values were also compared with the Total H3 values obtained by ChIP analysis for total histone H3 (Fig. 4F, upper panel). ChIPed DNA for the phosphorylated histone H3, acetylated histone H3, "total" H3, and methylated histone H3 were also quantitated using real-time PCR (Fig. 4, B, D, E, and G, respectively) with similar results.
Taken together, these results demonstrate that there is nuclear translocation of both p65 and p50 in response to LPS in both normal and tolerant cells, and that binding of p65 to the IL-1 promoter increases in LPS-responsive but not LPS-tolerant cells. p50 binding to the promoter is detected in the constitutive unstimulated state as well as in LPS-responsive and LPS-tolerant cells. Thus, p50 binding appears to remain unchanged throughout LPS treatment despite LPS-induced increases in nuclear p50 in tolerant cells (Fig. 3D). These in vivo results contrast our previous results using EMSA (26). We also found that LPS induced phosphorylation of histone H3 on Ser10 and coincident demethylation of histone H3 on Lys9 at the transcriptionally active IL-1 promoter. These modifications were not observed at the inactive IL-1 promoter in tolerant cells.
Discussion
NF-B regulation of transcription by the classical or canonical pathway can be divided into two phases (10, 29). First is a cytosolic phase in which there is IB kinase -dependent activation of IB followed by ubiquitin-directed IB degradation that allows translocation of p50 and p65 into the nucleus. The second or nuclear phase of NF-B regulation involves events that result in derepression NF-B-dependent promoters, chromatin remodeling, and binding of transcriptionally active p65 and p50 as a heterodimer to cognate DNA on enhancer and promoter elements. Assembly of the full promoter transcriptome follows this with activation of RNA polymerase II (30). This canonical mechanism for NF-B activation is fundamental in transactivating genes expressed by innate immune cells, and these gene products are critically involved in host protection and host injury, including septicemia (31). However, soon after septicemia is initiated, the NF-B pathway is disrupted at the level of transcription (4, 6, 16), and this occurs coincident with repression of proinflammatory genes such as TNF- and IL-1. Thus, these and other proinflammatory genes are reprogrammed during septicemia to be LPS tolerant. Surprisingly, there is clear evidence in circulating leukocytes and tissues showing chronic activation of NF-B with nuclear accumulation of p65 during sepsis (13, 15, 32), and these studies often use the inclusive and somewhat misleading term "activated NF-B."
To our knowledge we report the first observation that may explain the activation of classical pathway of cytosolic NF-B with accumulation of p65 in the nucleus, while there is repression of the nuclear phase of NF-B-dependent transcription. Using IL-1 as the NF-B-dependent gene and THP-1 cells as a model of septicemia leukocytes with LPS tolerance (the IL-1 gene is consistently repressed in human blood sepsis leukocytes, but variable in murine leukocytes), we show that the IL-1 promoter is disrupted, both at the level of p65 binding and nucleosome remodeling during LPS tolerance. LPS tolerance with repression of IL-1 transcription is a consistent feature of septicemia leukocytes studied ex vivo (4, 18).
We also provide the novel observation that p50 is bound to the IL-1 promoter in the basal, LPS-responsive, and LPS-tolerant states, without any observable changes. Our results support recent findings suggesting that p50 has evolved to constitutively bind to promoters of inflammatory genes without disturbing the nucleosomes or requiring their remodeling (33), thus providing rapid access to other transcription factors like p65. With pronounced increases in nuclear p50 levels in LPS-tolerant cells and the lack of any observable change in p50 binding to the IL-1 promoter, it is likely that the bulk of the p50 binding observed in LPS-tolerant cells occurs as a p50 homodimer, whereas the majority of the p50 found in LPS-responsive cells occurs as part of a p65-p50 heterodimer complex.
NF-B-dependent genes are usually controlled by p50 or p52 homodimers that repress transcription that depends on corepressors that bind to the homodimers (discussed in Ref. 34). It has been clearly shown that the p50 plays a role in LPS tolerance and that p50 homodimers predominate in the LPS-tolerant phenotype (14, 35, 36). One potential explanation for LPS tolerance at the level of the gene is formation of a repressor complex, which includes inhibitor corepressors and monomers or homodimers of p50, that are reassembled after the initial LPS response of septicemia stress (37, 38). We are currently exploring the possibility that other NF-B proteins and/or other interacting transcription factors may be assembled at the IL-1 promoter in LPS-tolerant cells.
Another level of regulation that may participate in transcription repression in LPS tolerance is chromatin remodeling of the promoter by histone modification (39). Our data show increased phosphorylation of Ser10 on histone H3 in the LPS-responsive phenotype and a hypophosphorylated state of this amino acid in the LPS-tolerant cell. Phosphorylation of H3 at Ser10 is a critical event in nucleosome remodeling, and our data further show that phosphorylation of H3 on Ser10 occurs in conjunction with demethylation of H3 on Lys9 (40). This methylation/phosphorylation shift supports the model that was recently offered as the hypothetical binary shift paradigm for chromatin remodeling (23). Although we did not observe demonstrable changes in acetylation of Lys9 or Lys14, it is possible that changes in acetylation occur at other residues or on other histone tails, or that histone acetylation is not required for regulating transcription factor binding to inflammatory genes. It is unclear whether activation of specific demethylases is a prerequisite to phosphorylation of H3 Ser10 or whether H3 Ser10 phosphorylation results in demethylation. Although nucleosome remodeling plays an important role in LPS tolerance at the IL-1 promoter, it is unlikely to be the only mechanism responsible for tolerance, since LPS tolerance disrupts transcription of both transiently and stably transfected reporter constructs (26). The higher order structure of transient transfected genes is distinct from the native chromatin organization (41, 42).
Of further potential importance are more intracellular signaling components such as kinases (43, 44, 45, 46) or phosphatases (47). p38 MAPK (43, 48), probably acting via another kinase such as mitogen-stimulated kinase and IB kinase (49, 50) are implicated in both phosphorylating p65 on Ser536 for enhanced transactivation, as well as phosphorylating histone H3 on Ser10 for nucleosome remodeling at or near the promoters of NF-B-dependent genes. Disruption of p38 kinase activity has been implicated in in vitro and in vivo models of LPS tolerance (reviewed in5), and others and we (data not shown) have found p38 inactive in LPS-tolerant THP-1 cells.
Clearly, the generation of an LPS-tolerant phenotype is complex and multifaceted. This report focuses on nuclear events at or near the promoter nucleosome of inflammatory genes. There is reprogramming of proximal cytosolic mediators such as IRAK monomyeloic, MyD88, Tollip, SHIP, single Ig IL-1R related protein, and perhaps suppressor of cytokine signaling 1 and 3 during LPS tolerance (51, 52, 53). The complexity of the LPS-tolerant phenotypes likely extends from the cell membrane into the nucleus, where the final pathways for repressing proinflammatory genes and sustaining the expression of anti-inflammatory genes operate. Fig. 5 is a schematic model of suppression of nuclear events involving NF-B at the IL-1 promoter of LPS-responsive vs LPS-tolerant THP-1 cells.
In summary, NF-B p65 binding and chromatin remodeling at the IL-1 promoter gene are differentially altered in the LPS-tolerant, when compared with the LPS-responsive, THP-1 monocytes. NF-B is activated in the cytosol of both LPS-responsive and LPS-tolerant cells with translocation of p65 and p50 into the nucleus. NF-B p65 binding to the IL-1 promoter is rapidly and transiently increased in LPS-responsive monocytes, whereas its binding to the IL-1 promoter is significantly limited in LPS-tolerant cells, an observation that correlates with repressed transcription of IL-1. In contrast, NF-B p50 binding is constitutive and is apparently unchanged by LPS stimulation in normal or tolerant cells. In LPS-responsive cells, there is also a binary shift in nucleosome remodeling with histone H3 demethylation of Lys9 and increased histone H3 phosphorylation of Ser10, when significant NF-B p65 binding is observed. However, in LPS-tolerant cells where NF-B p65 binding levels are reduced, there are concomitant reductions in phosphorylated histone H3 on Ser10 and steady levels of methylated histone H3 at Lys9 are retained around the IL-1 promoter. These results suggest that promoter-specific alterations contribute to the LPS-tolerant phenotype and may contribute to the suppressed innate immune response in septicemia. Analysis of promoter regulatory events of distinct genes during LPS tolerance might provide more targeted approaches to altering the disastrous outcome in septicemia.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by National Institutes of Health Grant AI-09169 (to C.E.M.) and American Lung Association Grants RG-035-N and RG-065-NL and American Heart Association Grant 0051211U (to B.K.Y.), and National Institutes of Allergy and Infectious Diseases AI-50089 (to L.L.).
2 Address correspondence and reprint requests to Dr. Barbara K. Yoza, Section of Molecular Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27157. E-mail address: byoza@wfubmc.edu
3 Abbreviation used in this paper: ChIP, chromatin immunoprecipitation.
Received for publication January 24, 2005. Accepted for publication April 12, 2005.
References
Angus, D. C., R. S. Wax. 2001. Epidemiology of sepsis: an update. Crit. Care Med. 29: S109-S116.
Bone, R. C., C. J. Grodzin, R. A. Balk. 1997. Sepsis: a new hypothesis for pathogenesis of the disease process. Chest 112: 235-243.
Beutler, B., K. Hoebe, X. Du, R. J. Ulevitch. 2003. How we detect microbes and respond to them: the Toll-like receptors and their transducers. J. Leukocyte Biol. 74: 479-485.
Cavaillon, J. M., C. Fitting, M. dib-Conquy. 2004. Mechanisms of immunodysregulation in sepsis. Contrib. Nephrol. 144: 76-93.
Fan, H., J. A. Cook. 2004. Molecular mechanisms of endotoxin tolerance. J. Endotoxin Res. 10: 71-84.
West, M. A., W. Heagy. 2002. Endotoxin tolerance: a review. Crit. Care Med. 30: S64-S73.
Cogswell, J. P., M. M. Godlevski, G. B. Wisely, W. C. Clay, L. M. Leesnitzer, J. P. Ways, J. G. Gray. 1994. NF-B regulates IL-1 transcription through a consensus NF-B binding site and a nonconsensus CRE-like site. J. Immunol. 153: 712-723.
Baldwin, A. S.. 2004. NFB: biology and pathology–Keystone Symposium. 11–16 January 2004, Snowbird, UT, USA. IDrugs 7: 215-216.
Baldwin, A. S.. 2001. Control of oncogenesis and cancer therapy resistance by the transcription factor NF-B. J. Clin. Invest. 107: 241-246.(Christopher Chan*, Liwu L)